Bias in the Use of Standard American Norms with Spanish Translations of the Wechsler Memory Scale-Revised

Abstract

Although most clinicians would readily agree that there is a need for Spanish translations and normative samples of major psychological tests because of the large number of individuals within the United States whose primary language is Spanish, there are in fact few well normed tests applicable to the Spanish-speaking client in the U.S. As a result, many clinicians administer cognitive tests normed on English-speaking American populations translated into Spanish, then interpret the test results using the standard American norms. The argument is generally made that such a procedure is a reasonable approximation and when cautiously interpreted can be useful. The present study investigated the impact of this practice by using the Wechsler Memory Scale-Revised (WMS-R) on Spanish-speaking adults. A group of 50 normal Hispanic Americans aged 25 to 34 years were administered the WMS-R translated into Spanish. Results showed that using English-language standard norms resulted in Spanish-speaking normal individuals getting scores an average of 1 standard deviation below "average." These findings were present not only for verbal but for nonverbal tests as well. Based on these results we strongly argue against the clinical practice of using translations of English language tests without renorming and running new validity tests.